biomass co-firing under oxy-fuel conditions: a computational fluid...

1

Biomass co-firing under oxy-fuel conditions: A computational fluid dynamics

modelling study and experimental validation

L. Álvarez1, C. Yin2, J. Riaza1, C. Pevida1, J.J. Pis1, F. Rubiera1*

1 Instituto Nacional del Carbón, INCAR-CSIC, Apartado 73, 33080 Oviedo, Spain

2 Department of Energy Technology, Aalborg University, 9220 Aalborg East, Denmark

Abstract

This paper presents an experimental and numerical study on co-firing olive waste (0,

10%, 20% on mass basis) with two coals in an entrained flow reactor under three oxy-

fuel conditions (21%O2/79%CO2, 30%O2/70%CO2 and 35%O2/65%CO2) and air-fuel

condition. Co-firing biomass with coal was found to have favorable synergy effects in

all the cases: it significantly improves the burnout while remarkably lowers NOx

emissions. The reduced peak temperatures during co-firing can also help to mitigate

deposition formation in real furnaces. Co-firing CO2-neutral biomass with coals under

oxy-fuel conditions can achieve a below-zero CO2 emission if the released CO2 is

captured and sequestered. The model-predicted burnout and gaseous emissions were

compared against the experimental results. A very good agreement was observed, the

differences in a range of ±5-10% of the experimental values, which indicates the model

can be used to aid in design and optimization of large-scale biomass co-firing under

oxy-fuel conditions.

Keywords: Biomass co-firing; Synergy effects; CO2 capture; Oxy-fuel combustion;

Below-zero CO2 emissions; CFD

* Corresponding Author: Tel : +34 985 119090; E-mail address: [email protected]. (F. Rubiera)

2

1. Introduction

The combustion of coal in power plants generates a large amount of CO2 which is one

of the major contributors to global climate change. A diverse power generation portfolio

including Carbon Capture and Storage (CCS) technologies and renewable energies is

needed to reduce current atmospheric CO2 concentration of 393 ppm to below 354 ppm

in 1990. During oxy-coal combustion, coal is burnt in a mixture of oxygen and recycled

flue gas (mainly CO2 and H2O), to yield a rich stream of CO2 which, after purification

and compression, is ready for sequestration [1]. Co-firing biomass with fossil fuels in

existing utility boilers is also a feasible technology which can not only significantly

reduce CO2 emissions but also increase the share of renewable energy sources in energy

systems [2, 3]. The combination of oxy-coal combustion with biomass co-firing could

afford a way to increase CO2 capture efficiency [4]. Biomass co-firing has been

successfully performed in over 200 installations worldwide for a large number of

combinations of fuels, either in pilot tests or as part of commercial enterprises [5].

Though problems may arise in relation to biomass transport costs and difficulties in

milling, these can be manageable if adequate consideration is given to the fuels, design

and operating conditions used in the burners and boilers [6].

Compared with biomass co-firing, oxy-fuel combustion is still in the demonstration

stage, with plans for industrial scale-up to come into effect in the near future [4]. The

successful implementation of oxy-fuel combustion depends on fully understanding the

difficulties that can arise from replacing nitrogen by CO2 in the oxidizer stream. Oxy-

fuel conditions strongly promote radiative heat transfer, as a result of the much higher

levels of CO2, H2O, and in-flame soot, as well as the different CO2/H2O ratio to that of

air-firing combustion [7]. Other aspects of combustion, such as volatile combustion,

flame ignition and stability, or pollutant formation, may also be affected [8]. Biomass

3

co-firing under oxy-fuel is likely to bring up more uncertainties. Biomass co-firing

under oxy-fuel conditions is an attractive option to simultaneously increase the use of

renewable energy sources, exploit the favorable synergy effects of biomass/coal co-

firing and achieve below-zero CO2 emissions, which has been very little investigated so

far [9-12]. For instance Smart et al. [11] evaluated the impact of co-firing biomass on

pollutant formation, burnout and heat transfer under oxy-fuel conditions in a 0.5 MWt

combustion test facility. Experiments at laboratory scale have been focused on the

effects of different co-firing ratios on burnout and NO emissions [12].

Computational Fluid Dynamics (CFD) models have been used to simulate pulverised

coal and biomass co-firing in conventional combustion systems [13-15]. In recent years,

on the basis of the accumulated knowledge of the fundamental differences between air-

fuel and oxy-fuel combustion, much effort has been devoted to developing and

validating sub-models for the new combustion environment. For instance, new

approaches have been developed for heat transfer modelling in environments with high

concentrations of CO2 and H2O vapor, e.g., the Weighted-Sum-of-Gray-Gases-Model

(WSGGM) refined for oxy-fuel combustion modelling [16, 17]. Specific models for

volatile combustion in CO2–rich environments [18] and for char combustion under oxy-

fuel conditions [19] have also been developed. In terms of modelling of biomass co-

firing under oxy-fuel conditions, a recent paper by Holtmeyer at al. [20] could be the

only effort in literature, in which the effect of co-firing sawdust and a subbituminous

coal on NO emissions in air and oxy-fuel conditions was studied.

This paper is to comprehensively study co-firing biomass with coals under air-fuel and

oxy-fuel conditions in a lab-scale entrained flow reactor (EFR). The computational fluid

dynamics (CFD) study is the main workhorse and the experimental study is used to

validate the CFD modelling. Among others, the effects of biomass shares in co-firing

4

(from 0 to 20 wt% of biomass), firing conditions, and different coals to be separately

used in co-firing on the overall combustion behavior and NO emissions are investigated

and discussed.

2. Experimental setup

All the combustion experiments of this study were performed in a down-fired EFR,

which has been introduced in detail in Riaza et al. [21]. Figure 1 shows a schematic

diagram of the reactor. The EFR has an internal diameter of 4 cm and a length of 200

cm. For the experiments reported in this work a reaction zone of 140 cm was used. The

EFR was electrically heated and the preheated gases were introduced through flow

straighteners to ensure laminar flow conditions. The experiments were performed at a

heated furnace temperature of 1273 K. The gas flow was set to 22.4 L/min, which

corresponds to a residence time of 2.5 s. The amount of excess oxygen in the oxidant

over the required stoichiometric oxygen was set to %25,2 =exO . The excess oxygen was

then used to calculate the required fuel mass flow rate, )1/( ex,2st, Omm FF += && , where

st,Fm& represents the stoichiometric fuel mass flow rate. Four different combustion

atmospheres were employed: air (21%O2/79%N2), and three binary gas mixtures of O2

and CO2 (21%O2/79%CO2, 30%O2/70%CO2 and 35%O2/65%CO2).

- Figure 1 here -

Two coals of different rank were used in this work: a semi-anthracite from the Hullera

Vasco-Leonesa in León, Spain (HVN); and a South African high-volatile bituminous

coal from the Aboño power plant (350 MWe) in Asturias, Spain (SAB). A biomass,

olive waste (OW) was also employed. This biomass is the solid waste that remains after

the process of pressing and extraction of olive oil. The coal and biomass samples were

ground and sieved to obtain a particle size fraction of 75-150 μm. The proximate and

5

ultimate analyses together with the high heating values of the fuel samples are presented

in Table 1.

- Table 1 here -

The fuel samples in a hopper were fed into the EFR reaction zone through a cooled

injector. The fuel particles were then injected into the centre of the preheated gas

stream. The reaction products were quenched by aspiration in a stream of nitrogen by

means of a water-cooled probe. The particles were removed by means of a cyclone and

a filter, and the coal burnout was determined by the ash tracer method. The exhaust

gases were monitored using a battery of analyzers Emerson X-Stream X2GP with non-

dispersive infrared photometers detectors for CO, CO2, SO2, and NO; and a

paramagnetic sensor for O2.

The experimental findings on burnout (defined in this study as the ratio of mass loss of

a fuel sample during its combustion to the original mass of the fuel feed) and NO

emissions have been reported in a previous paper on experimental study [12]. Here, the

results were used to validate the CFD simulations, in order to evaluate the effect of

biomass co-firing on burnout and NO emissions under various conditions and to

establish the modelling capability for oxy-fuel combustion of biomass/coal blends.

3. Numerical modelling

When solid fuel particles are injected into the reactor, some interdependent processes

(e.g., gas and particle phase dynamics, turbulence, heat transfer, pollutants formation,

and homogenous and heterogeneous reactions) take place and need to be appropriately

taken into account in modelling. In this study, the simulations were performed using a

commercial CFD package, Ansys Fluent version 13 [22]. Most of the principal

equations of the combustion sub-models were explained in detail in the Fluent Theory

Guide. Here, only the user-defined submodels (i.e., for oxy-coal radiation) or the

6

significant reactions and their kinetics which are very dependent on fuel and operation

conditions (i.e., char and volatile combustion or fuel devolatilisation), are described in

detail. Details of the mesh and boundary conditions are also given.

Because of the symmetrical conditions at two perpendicular mid-planes, only a quarter

of the total reaction zone of the EFR (4 cm i.d., 140 cm height) was used in the

simulations. The simulation domain was meshed into a high-quality structured grid

consisting of about 75,000 hexahedral cells, which was same as that used in our

previous oxy-coal combustion study [23]. The mesh was found fine enough to assure

grid-independent simulation results. Totally 24 different combustion tests were

performed in our experimental study, i.e., six different fuel samples or blends under four

atmospheres, respectively. Table 2 shows in detail the share of biomass (on mass basis)

in the fuel blend, feeding rate of the blend, composition and mass flowrate of the

gaseous reactant into the reactor for each of the test cases. The temperature of the

reactor wall and the injector wall was 1273 K and 373 K, respectively. All the 24 test

cases were numerically simulated here.

- Table 2 here -

3.1. Gas and particle phase dynamics

Transport equations for the continuous phase were solved using an Eulerian approach

and the fuel particles were tracked in a Lagrangian frame of reference through the

calculated gas field. The trajectories of solid fuel particles were computed using the

discrete phase model, by assuming spherical particles and retaining only the drag and

gravity forces in the equation of motion of the particles. The effect of turbulence was

accounted for by the RNG k-ε turbulence model. Although this model was developed

and mainly used for high-Reynolds number turbulent flows, the use of a differential

equation for turbulent viscosity, which is derived from the scale elimination procedure

7

in RNG theory, also enables the model to better and successfully handle low-Reynolds

number and near-wall flows [22, 24, 25]. The size distribution of the coal particles was

fitted to a Rosin-Rammler distribution. The minimum, mean and maximum particle

diameters are 75, 115 and 150 µm, respectively. The spread parameter is unity and 5

size groups are considered. The effect of turbulence on particle dispersion was taken

into account using the discrete random walk model. Totally 2400 particle streams are

tracked in each case.

Here it has to be mentioned that in this study both the tracking and conversion of

biomass particles were modelled in the same methodology as coal particles. In general,

biomass particles are often fibrous and non-friable. Therefore, biomass particles

prepared and used for suspension co-firing are often much larger in size and more

irregular in shape than pulverized coal particles. Yin et al. [13] developed a model to

track large, nonspherical particles in fluid flows, which was validated using their

experimental results of the motion of a large cylindrical particle (length of 50 mm and

aspect ratio of about 9) in a water flow and then applied to biomass/natural gas co-firing

modelling. They also extended their efforts to large biomass particle conversion by

accounting for various intra-particle processes in modelling and found for pulverized

biomass particles of a few hundred microns in diameters the intra-particle heat and mass

transfer was a secondary issue at most in particle conversion under suspension-firing

conditions [26, 27]. Gubba et al. [28] developed a model to evaluate the influence of

particle shape and internal thermal gradients on biomass suspension co-firing flames. In

this study, the biomass particles are tiny in size (below 200 μm) and near-spherical in

shape. Under the conditions of all the test cases, the Biot number is well below 0.1, i.e.,

1.0/Bi 61 <<⋅≡ kdh p , where h , pd and k represent convection heat transfer coefficient

between the gas and solid fuel particle, diameter of solid fuel particle, and heat

8

conductivity of solid fuel particle, respectively. So the particles can be considered to be

under isothermal conditions. This is also in line with the conclusion that isothermal

particle assumption may be no longer valid when the pulverised particle size exceeds

150-200 μm [29]. The biomass particles are assumed to be spherical, so only the gravity

and the standard drag force were retained in the equation of motion for biomass

particles, which is the same way used for coal particle tracking in this study.

3.2. Coal and biomass devolatilisation

In this study the devolatilisation rate of the solid fuels was modelled using a single step

first-order Arrhenius equation. For both the coals, the pre-exponential factor (A) and

activation energy (Ea) were obtained by means of the FG-DVC (Functional Group-

Depolymerisation Vaporation Crosslinking) code [30], and the values were estimated

assuming a final temperature of 1273 K and a heating rate of 105 K/s. For the biomass,

the kinetic values of A and Ea were obtained from an extensive review paper on biomass

pyrolysis [31]. Table 3 summarizes the kinetic data used as the devolatilisation model

inputs in the modelling study.

- Table 3 here -

It is necessary to point out that the FG-DVC predictions were determined for an inert

atmosphere. The release rate of the volatiles was not remarkably affected by the high

CO2 concentrations [32], although in the later stages of combustion there may be a

certain degree of char gasification with CO2. In a previous work of Álvarez et al. [33],

comparison of coal devolatilisation in N2 and CO2 atmospheres was conducted in a

thermogravimetric apparatus. Before devolatilisation was finished at around 1150 K, the

mass loss curves in the N2 and CO2 atmospheres followed a similar trend. The

additional mass loss in the CO2 atmosphere above 1173 K, when the release of volatiles

had already finished, was mainly due to the char-CO2 reaction. Thus, we employed the

9

same devolatilisation kinetics for the air- and oxy-firing CFD simulations. Then a char-

CO2 reaction was included in the char combustion sub-model.

3.3. Volatile combustion

During the different conversion processes of fuel particles along their trajectories, some

species are released into gas phase (e.g., volatiles and CO), creating sources for gas

phase combustion. The volatiles often carry a large percentage of the energy of solid

fuels (e.g., about 50% for coals and an overwhelming majority for biomass). The

homogeneous combustion of the volatiles also plays a vital role in ignition and flame

stability, local temperature, species distribution and pollutant formation. Moreover, high

CO2 concentration under oxy-fuel conditions may also affect the gas-phase combustion.

Therefore, gas-phase combustion mechanism is expected to play an important role in

modelling of the co-firing processes. To describe the gas composition in the EFR, the

species transport approach was used, with the Eddy-Dissipation Concept for the

turbulence-chemistry interaction. In the CFD simulations, the volatile gases were

lumped into one single “artificial” species, CHyOx. The compositions and the formation

enthalpies were determined from the proximate and ultimate analyses of the fuels. The

lumped volatiles were represented by CH3.87O0.11, CH3.56O0.47 and CH2.45O0.93, and their

formation enthalpies (in J/kmol) were -4.2×107, -5.4×108 and -4.0×108 for the HVN,

SAB and OW, respectively. For the air combustion cases, six species were defined as

follows: CHyOx, O2, H2O, CO, CO2 and N2, and the original Jones and Lindstedt 4-step

(JL-4) global mechanism [34], was employed:

CHyOx + (0.5+0.25y –0.5x)O2 → CO + 0.5y H2O (R1)

CO+ 0.5 O2 → CO2 (R2)

CO + H2O → CO2 + H2 (R3)

H2 + 0.5 O2 → H2O (R4)

10

For oxy-coal combustion cases, CO2 was set as the last species and a refined JL-4

mechanism with the kinetics adapted for oxy-fuel combustion conditions was employed

[18]. In comparison with the original JL-4 step mechanism, the refined JL-4 mechanism

retains the initiating reactions involving hydrocarbon and O2 (R1), whilst refines the

CO-CO2 reactions (R2) in order to improve prediction of major species concentration.

The refined JL 4-step mechanism was also implemented in CFD modelling of a 0.8 MW

oxy-natural gas flame, and both the predicted gas species (e.g., CO2, O2, CO and H2)

and gas temperature showed a good agreement with experimental results [35].

3.4. Char oxidation

For char oxidation, the multiple-surface-reaction char model was employed. At the

temperatures of this EFR, CO is the dominant product in char oxidation C(s) +O2 →

CO/CO2. Therefore, only the reaction (R5) was considered in char oxidation modelling.

The CO formed in (R5) will undergo further reactions as shown in (R2) and (R3).

C(s) + 0.5 O2 → CO (R5)

Global kinetics for the coal char combustion rates, represented as Arrhenius

expressions, were based on the activation energies and pre-exponential factors as

determined by Gil et al. [36], as summarized in Table 4.

- Table 4 here -

At the temperature of this work (1273 K), combustion takes place in Regime II (kinetics

and diffusion control). The effectiveness factor (η), defined as the ratio of the diffusion

rate to the maximum diffusion rate, is a measure of the extent of penetration of the

oxidant into the char matrix. It was determined via the Thiele modulus as calculated by

Gharebaghi et al. [19] for a large number of defined oxy-fuel combustion cases. At

elevated temperature and CO2 concentrations, char CO2 gasification could become more

important particularly in the later stages of the oxy-fuel combustion process. As long as

11

the oxygen concentration is higher, the contribution of the gasification reactions will be

much smaller than that of the oxidation reactions. Nevertheless, char gasification was

also considered in the modelling study (R6), and its kinetic data was obtained from the

literature [37, 38]. In the current simulations, the effect of CO2 gasification reaction

accounted for less than 1% on coal burnout.

C(s) + CO2 → 2CO (R6)

The combustion of biomass char is more complicated since it is affected not only by the

composition of the biomass, but also by the shape and size of the particle. Different

approaches can be found in literature, such as the diffusion-limited surface reaction rate

model modified by the aspect ratio-dependent enhancement factor [13], or Smith’s

intrinsic model modified by a constant enhancement of 4 to represent the higher burning

rate of the biomass char particles [14]. In this study, both the raw biomass particles and

the resulting biomass char particles are near-spherical and the carbon content in the

biomass is relatively low (about 20%). Therefore, no special enhancement factor was

used for biomass char oxidation. The kinetics for biomass char combustion were

obtained from [39, 40].

3.5. Heat transfer model

In a pulverised fuel chamber, radiation is often the dominant mode of heat transfer. The

most widely used model for gaseous radiative properties is the Weighted-Sum-of-Gray-

Gases-Model (WSGGM), which represents a reasonable compromise between an

oversimplified gray gas model and a comprehensive approach addressing high-

resolution dependency of radiative properties and intensity upon wavelength. The model

parameters derived by Smith et al. [41] for several partial pressures of CO2 and H2O

vapor in typical air-fuel combustion are often used in combustion modelling. The

original Smith et al. WSGGM is a one-clear-gas, three-gray-gas model. However, it is

12

often implemented in a further simplified way in which constant properties over the

entire spectrum are assumed and only one single radiative transfer equation per

direction is solved for the entire spectrum, e.g., as implemented in Ansys Fluent.

Recently, some efforts are made to refine and extend the WSGGM for oxy-fuel

conditions. For instance, Yin et al. [16] used the exponential wide band model, the same

reference model as Smith et al. used in their derivation, to generate a much broader

emissivity database covering also oxy-fuel conditions, and derived new WSGGM

parameters by using an improved data-fitting technique. The refined WSGGM also

included more representative conditions to better account for the variations in H2O/CO2

molar ratio in an oxy-fuel flame. The importance of nongray-gas effects in modelling of

large-scale oxy-fuel combustion was also well demonstrated and nongray calculation of

an appropriate WSGGM was highly recommended for combustion modelling [42].

In this study, the Discrete Ordinates model was employed for radiative transfer equation

for both oxy- and air-fuel combustion. In the oxy-fuel combustion modelling, non-gray

calculation of the oxy-fuel WSGGM [16] was performed to evaluate the gaseous

radiative properties. Compared to the conventional gray calculation of the Smith et al.

air-fuel WSGGM, it does not remarkably compromise computational efficiency while

has inherent potentials to improve simulations of oxy-fuel combustion processes, in

which the degree of improvement depends on the scale or the beam length of the

furnace under study. For the simulations in the oxy-fuel atmosphere with the highest

CO2 content (i.e., 21% O2/79% CO2) non-gray oxyfuel WSGGM and gray air-fuel

WSGGM calculations were both performed. Since the reactor under simulation is small

(in terms of beam length), using new gaseous radiation properties led to little

differences in CFD predictions, just as expected. The deviations in temperature profiles

for both cases were less than ±1%.

13

3.6. NOx formation

NOX simulations were carried out as a post-processor. For air-firing conditions, both the

thermal and fuel-NO were considered. For oxy-firing conditions, fuel-NO formation

was considered to be the dominant mechanism.

Fuel-NO is formed from the oxidation of molecular nitrogen organically bound within

the fuel mass. Fuel-bound nitrogen can either be released during devolatilisation

(referred to as volatile-N), or can remain in the char (referred to as char-N). The

partitioning of nitrogen between char and volatiles, summarized in Table 5, was

determined experimentally. Usually, it is assumed that HCN and NH3 are the dominant

species formed as nitrogen-bearing intermediates at rates that depend both on the local

combustion conditions and on the nitrogen content of each specific fuel [43]. For

biomass, it was assumed that NH3 was the only nitrogen intermediate, whereas for coals

the HCN and NH3 release rates were based on the prediction made by Álvarez et al. [44]

HCN and NH3 are competitively oxidised and reduced to form NO and N2, respectively,

according to DeSoete’s scheme:

HCN + O2 → NO + … RTOHCN eXXR /95.280451

210101 −⋅⋅⋅⋅= (1/s) (R7)

HCN + NO → N2 + ... RTNOHCN eXXR /25115112103 −⋅⋅⋅⋅= (1/s) (R8)

NH3 + O2 → NO + ... RTONH eXXR /2.133947

236104 −⋅⋅⋅⋅= (1/s) (R9)

NH3 + NO → N2 + ... RTNONH eXXR /95.113017

38108.1 −⋅⋅⋅⋅= (1/s) (R10)

where HCNR& , 3NHR& , X , uR and T represent the conversion rate of HCN (1/s),

conversion rate of NH3 (1/s), molar fraction, universal gas constant (=8.315 J/(mol·K)),

and gas temperature (K), respectively.

The fuel nitrogen partitioning between the char and the volatiles can be estimated by

pyrolysis network codes, obtained from the literature or determined experimentally. In

this work, the partitioning of the fuel-bound nitrogen between the char and volatiles was

14

determined experimentally during the devolatilisation of the samples in N2 and CO2 in

the EFR at 1273 K. From the analysis of N content of the parent samples and that of the

chars, and using the ash as a tracer method, the char and char-N yields can be

determined and, thus, the nitrogen content in char and volatiles was estimated. Char-N

was assumed to be heterogeneously oxidised to NO, for both coal and biomass char

particles, with an assumed conversion factor of 20% (mass/mass %) [45].

- Table 5 here –

4. Results and discussion

4.1. Overall combustion behaviour

In biomass combustion, the higher volatile yields produce more off-gas, which causes

the off-gas to proceed to a much larger volume in the reactor before mixing with the

oxidizer and completing gas-phase combustion. Therefore, biomass flame tends to

occupy a much larger flame volume (i.e., the regions with low O2 and high CO

concentration) than coal flame [27, 28]. The large flame volume effectively evens out

the temperature distribution and lowers the peak temperatures, which favors a lower

pollutant formation (e.g., lower NOx emission) and a reduced deposition potential. The

comparatively low (relative to the peak temperature in a coal flame, still high enough)

and uniform temperature distribution in a much larger flame volume also favors a

higher burnout. Both improved burnout and decreased NOx emissions were observed in

this study when the coals were co-fired with biomass under different operating mode.

Table 6 shows a comparison between the experimental and predicted burnouts. A very

good agreement can be observed, in which the difference between the experimental

results and model predictions fell within a range of ±5-10% of the experimental values

for all the test cases.

- Table 6 here -

15

As can be seen from Table 6, co-firing biomass led to a noticeable increase in the

burnout for both coals. This effect became more pronounced as the biomass share

increased, especially in the atmospheres with a lower O2 content, i.e., 21%O2/79%N2

and 21%O2/79%CO2. For the HVN-OW blends, there was a great improvement in the

burnout, especially when switching from 10% to 20 wt% of biomass addition.

Comparatively, the improvement was not so obvious for the SAB-OW blends, since the

individual coal SAB had already reached a high burnout degree before being blended

with the biomass and there was less margin for improvement. Table 6 also shows the

predicted burnouts very well reproduce the trend observed in experiments, indicating

that the CFD model successfully describes the co-firing behaviour in the air and oxy-

fuel conditions.

More details about the co-firing characteristics under various conditions in the EFR can

be revealed from the CFD modelling study. Figure 2 shows the temperature contours in

the mid plane of the EFR for both its whole length and its first 40 cm, during HVN-OW

co-firing under different conditions, whereas Figure 3 shows the local fluctuations in the

area-weighted average temperature profiles around the location where the solid fuel

stream is ignited. In all cases the devolatilisation of the fuels takes place after the

injection of the fuel (at a distance of 0.15 m from the top of the reactor) and there is an

increase in temperature due to the heat released during the combustion of the volatiles.

Significant differences can be observed between the 21%O2/79%N2 and the

21%O2/79%CO2 atmosphere. When N2 is replaced by CO2, the gas temperatures drop

significantly due to the higher specific molar heat of CO2. It can also be observed that to

obtain similar gas temperatures in oxy-firing conditions to those of air-firing, the

oxygen content in the CO2 mixture must be of the order of ~30-35% to counteract the

negative effect of the higher CO2 heat capacity. Increasing the O2 percentage in CO2 up

16

to 30 or 35% is still insufficient to match the heat capacity of the air. However, under

the 30%O2/70%CO2 and 35%O2/65%CO2 atmospheres, the increase in the mass flux

rate (cf. Table 2) promotes the consumption rate of the volatiles, providing extra heat

feedback to the fuel particles and leading to higher gas and particle temperatures.

- Figs. 2 and 3 here -

Figure 2 shows the temperature distribution in the EFR. As expected, the addition of

biomass made some difference in the gas temperature distribution. Co-firing biomass

with coal tends to even out the high-temperature zones into a larger volume but with

decreased peak temperatures, especially under oxy-fuel conditions, which leads to a

higher burnout and a lower emission. This observation is in line with the findings in

literature. For instance, Molcan et al. [46] carried out biomass/coal co-firing

experiments in a 3 MWth combustion test facility, and found that biomass addition to

coal not only improved combustion efficiency but also led to lower flame temperatures.

Yin et al. [28] performed simulations for coal and straw combustion in a swirl-stabilized

burner, and in the case of the straw-flame the gas temperature predictions were slightly

lower than for the coal-flame. Ma et el. [14] also performed a numerical study for co-

firing coal and biomass, and their predictions (as well as the experimental values)

showed that the addition of biomass (wood, Miscanthus and olive waste) to coal would

result in a decrease in gas temperatures.

Figure 4 shows the counterpart for SAB-OW co-firing under various conditions: the

temperature contours in the mid plane of the EFR for both its whole length and its first

40 cm, whereas Figure 5 shows the evolution of the area-weighted temperature of the

gases with axial location. The similar tendency as in HVN-OW co-firing can also be

observed here.


17

Figures 6 and 7 show the burning rates in air and oxy-firing conditions during the HVN-

OW co-firing and SAB-OW co-firing, respectively. For both blends, differences in

combustion behaviour under air and oxy-fuel conditions are apparent. When the co-

firing condition is switched from air-fuel (i.e., 21%O2/79%N2) to low O2 concentration

oxy-fuel (i.e., 21%O2/79%CO2), the overall gas temperature level drops significantly,

which leads to a reduced burnout (as seen in Table 6). When increasing O2

concentrations in the oxy-fuel conditions (e.g., 30%O2/70%CO2 and 35%O2/65%CO2),

an increase in the burning rate is observed, because of the increased temperature levels

and the increased oxidizer concentration. This, in turn, gives rise to a higher burnout, as

shown in Table 6.


4.2. NOx emissions

Table 6 shows a comparison between the experimental and predicted NO emissions for

all the test cases. A very good agreement between the experimental and predicted results

is achieved.

SAB and HVN have very similar nitrogen contents: 1.8% vs. 1.7% on dry basis.

However, the burnout in SAB/OW co-firing is obviously higher than HVN/OW co-

firing for all the OW shares (from 0 to 20% on mass basis), meaning a higher char-N

conversion to NO in the former. As a result, SAB-OW co-firing is expected to produce

higher NO emission than HVN-OW co-firing. However, NO emission from SAB-OW

co-firing was found to be lower than that from HVN-OW co-firing in almost all the test

cases. One of the explanations is the much higher volatiles in SAB, which are released

and burned in the lower part of the EFR and serve as a re-burning agent to reduce the

formed NO to N2. The higher amount of NH3 and HCN released may also favour the

reduction of the formed NO to N2 via reactions (R8) and (R10).

18

For both the coals, NO concentration decreases when they are co-fired with biomass in

all the test conditions, even though the biomass has similar nitrogen content than both

the coals. The reduced peak temperatures during co-firing are one of the reasons for the

decreased NO emission. The large amount of off-gas released from the biomass may

proceed to a larger volume in the EFR prior to mixing with oxidizer and burnout. The

reducing atmosphere in part of the reactor favours the reduction of the NO formed to

N2. Comparatively, the decrease in NO emission from SAB-OW co-firing is more

remarkable than that from HVN-OW co-firing, thanks to the much higher volatile

content in SAB.


Figures 8 and 9 show more details of the predicted NO concentration profiles in the mid

plane of the EFR during HVN-OW co-firing and SAB-OW co-firing, respectively. A

decrease in NO emissions is observed when switching from conventional air-firing

condition to 21%O2/79%CO2 oxy-firing condition, mainly because of the decreased

temperature levels. NO emissions will increase monotonously as O2 concentrations in

the oxy-firing conditions increases (e.g., to 30% or 35%).

5. Conclusions

A comprehensive experimental and numerical study on co-firing olive waste with two

coals of different rank in a laboratory scale entrained flow reactor at 1273 K under

conventional air-fuel condition and three oxyfuel conditions was undertaken. Favorable

synergy effects were observed in the co-firing tests: co-firing biomass with coal not

only improved burnout but also reduced NOx emissions. The CFD-predicted burnout

and gaseous emissions were compared against the experimental results. A very good

agreement was achieved for all the cases: the differences between experimental and

modelling results fell in a range of ±5-10% of the experimental values, indicating a

19

good potential of the modeling routine for large-scale biomass co-firing under oxy-fuel

conditions.

Acknowledgements

L.A. acknowledges funding from the CSIC JAE program, co-financed by the European

Social Fund. J.R. acknowledges funding from the Government of the Principado de

Asturias (Severo Ochoa program). Financial support from the CSIC (Project PIE

201080E09) is gratefully acknowledged.

References

1. T. Wall, Y. Liu, C. Spero, L. Elliott, S. Khare, R. Rathnam, F. Zeenathal, B.

Moghtaderi, B. Buhre, C. Seng, R. Gupta, T. Yamada, K. Makino, J. Yu, An

overview on oxyfuel coal combustion-State of the art research and technology

development, Chemical Engineering Research and Design 87 (2009) 1003-1016.

2. A. Williams, M. Pourkashanian, J.M. Jones, Combustion of pulverised coal and

biomass, Progress in Energy and Combustion Science 27 (2001) 587-610.

3. A. Demirbas, Combustion characteristics of different biomass fuels, Progress in

Energy and Combustion Science 30 (2004) 219-230.

4. T.F. Wall, R. Stanger, S. Santos, Demonstrations of coal-fired oxy-fuel technology

for carbon capture and storage and issues with commercial deployment, International

Journal of Greenhouse Gases Control 5S (2011) S5-15.

5. IEA Bioenergy Task 32: Biomass combustion and co-firing, database of biomass co-

firing. http://www.ieabcc.nl/ (last visit November 2013).

6. S. Van Loo, J. Koppenjan, The handbook of biomass combustion and co-firing

(2008)

20

7. J. Andersson, R. Johansson, S. Hjärtstam, F. Johnsson, B. Leckner, Radiation

intensity of lignite oxy-fuel flame, Experimental Thermal Fluid Science 33 (2008)

67-76.

8. L. Chen, S.X. Yong, A.F. Ghoniem, 2012. Oxy-fuel combustion of pulverized coal.

Characterization, fundamentals, stabilization and CFD modelling, Progress in Energy

and Combustion Science 38 (2012) 156-214.

9. S. Black, J. Szuhánszki, A. Pranzitelli, L. Ma, P.J. Stanger, D. B. Ingham, M.

Pourkashanian, Effects of firing coal and biomass under oxy-fuel conditions in a

power plant boiler using CFD modelling, Fuel 113 (2013) 780-786.

10. J. Szuhánszki, S. Black, A. Pranzitelli, L. Ma, P.J. Stanger, D.B. Ingham, M.

Pourkashanian, Evaluation of the performance of a power plant boiler firing coal,

biomass and a blend under oxy-fuel conditions as a CO2 capture technology, Energy

Procedia 37 (2013) 1413-1422.

11. J.P. Smart, R. Patel, G.S. Riley, Oxy-fuel combustion of coal and biomass, the effect

on radiative and convective heat transfer and burnout, Combustion and Flame 157

(2010) 2230-2240.

12. J. Riaza, M.V. Gil, L. Álvarez, C. Pevida, J.J. Pis, F. Rubiera, Oxy-fuel combustion

of coal and biomass blends, Energy 41 (2012) 429-435.

13. C. Yin, L. Rosendahl, S.K. Kær, T.J. Condra, Use of numerical modeling in design

for co-firing biomass in wall-fired burners, Chemical Engineering Science 59 (2004)

3281-3292.

14. L. Ma, M. Gharebaghi, R. Porter, M. Pourkashanian, J.M. Jones, A. Williams,

Modelling methods for co-fired pulverised fuel furnaces, Fuel 88 (2009) 2448-2454.

15. M. Agraniotis, N. Nikolopoulos, A. Nikolopoulos, P. Grammelis, E. Kakaras,

Numerical investigation of Solid Recovered Fuels’ co-firing with brown coal in large

21

scale boilers – Evaluation of different co-combustion modes, Fuel 89 (2010) 3693-

3709.

16. C. Yin, L.C.R. Johansen, L.A. Rosendahl, S.K. Kær, New weighted sum of gray

gases model applicable to computational fluid dynamics (CFD) modeling of oxy-fuel

combustion: Derivation, validation, and implementation, Energy and Fuels 24 (2010)

6275-6282.

17. R. Johansson, B. Leckner, K. Andersson, F. Johnsson, Account for variations in the

H2O to CO2 molar ratio when modelling gaseous radiative heat transfer with the

weighted-sum-of-grey-gases model, Combustion and Flame 158 (2011) 893-901.

18. J. Andersen, C.L. Rasmussen, T. Giselsson, P. Glarborg, Global combustion

mechanisms for use in CFD modeling under oxy-fuel conditions, Energy and Fuels

23 (2009) 1379-1389.

19. M. Gharebaghi, R.M. Irons, M. Pourkashanian, A. Williams, An investigation into a

carbon burnout kinetic model for oxy-coal combustion, Fuel Processing Technology

92 (2011) 2455-2464.

20. M.L. Holtmeyer, B.M. Kumfer, R.L. Axelbaum, Effects of biomass particle size

during cofiring under air-fired and oxyfuel conditions, Applied Energy 93 (2012)

606-613.

21. J. Riaza, L. Álvarez, M.V. Gil, C. Pevida, F. Rubiera, J.J Pis, Effect of oxy-fuel

combustion with steam addition on coal ignition and burnout in an entrained flow

reactor, Energy 36 (2011) 5314-5319.

22. ANSYS FLUENT, 2011. Ansys Inc. USA

23. L. Álvarez, M. Gharebaghi, M. Pourkashanian, A. Williams, J. Riaza, C. Pevida, J.J.

Pis, F. Rubiera, CFD modelling of oxy-coal combustion in an entrained flow reactor,

Fuel Processing Technology 92 (2011) 1489-1497.

22

24. W.P. Jones, B.E. Launder, The prediction of laminarization with a two-equation

model of turbulence, International Journal of Heat and Mass Transfer 15 (1972) 301-

314.

25. S.A. Orszag, V. Yakhout, W.S. Flannery, F. Boysan, D. Choudhury, J. Maruzewski,

B. Patel, Renormalization group modelling and turbulence simulations, In

International Conference on Near-Wall Turbulence Flows, 1993. Tempe, Arizona.

26. C. Yin, S.K. Kær, L. Rosendahl, S.L. Hvid, Co-firing straw with coal in a swirl-

stabilized dual-feed burner: Modelling and experimental validation, Bioresource.

Technology 101 (2010) 4169-4178.

27. C. Yin, L. Rossendahl, S.K. Kær, Towards a better understanding of biomass

suspension co-firing impacts via investigating a coal flame and a biomass flame in a

swirl-stabilized burner flow reactor under same conditions, Fuel Processing

Technology 98 (2012) 65-73.

28. S.R. Gubba, L. Ma, M. Pourkashanian, A. Williams, Influence of particle shape and

internal thermal gradients of biomass particles on pulverised coal/biomass co-fired

flames, Fuel Processing Technology 92 (2011) 2185-2195.

29. Y.B. Yang, V.N. Sharifi, J. Swithenbank, L. Ma, L. Darwell, J.M. Jones, M.

Pourkashanian, A. Williams, Combustion of a single particle of biomass, Energy and

Fuels 22 (2008) 306-316.

30. P.R. Solomon, T.H. Fletcher, Impact of coal pyrolysis on combustion, Symposium

(International) on Combustion 15 (1994) 463-474.

31. C. Di Blasi, Modeling chemistry and physical processes of wood and biomass

pyrolysis, Progress in Energy and Combustion Science 34 (2008) 47-90.

23

32. R.K. Rathman, L.K. Elliot, T.F. Wall, Y. Liu, B. Moghtaderi, Differences in

reactivity of pulverised coal in air (O2/N2) and oxy-fuel conditions (O2/CO2), Fuel

Processing Technology 90 (2009) 797-802.

33. L. Álvarez, C. Yin, J. Riaza, C. Pevida, J.J. Pis, F. Rubiera. Oxy-coal combustion in

an entrained flow reactor: Application of specific char and volatile combustion and

radiation models for oxy-firing conditions, Energy 62 (2013) 255-268.

34. W.P. Jones, R.P. Lindstedt, Global reaction schemes for hydrocarbon combustion,

Combustion and Flame 73 (1988) 233-249.

35. C. Yin, L.A. Rosendahl, S.K. Kær, Chemistry and radiation in oxy-fuel combustion.

A computational fluid dynamics modelling study, Fuel 90 (2011) 2519-2529.

36. M.V. Gil, J. Riaza, L. Álvarez, C. Pevida, J.J. Pis, F. Rubiera, Oxy-fuel combustion

kinetics and morphology of coal chars obtained in N2 and CO2 atmospheres in an

entrained flow reactor, Applied Energy 91 (2012) 67-74.

37. G.-S. Liu, S. Niksa, Coal conversion submodels for design applications at elevated

pressures. Part II. Char gasification, Progress in Energy and Combustion Science 30

(2004) 679-717.

38. J. Fermoso, C. Stevanov, B. Moghtaderi, A. Arias, C. Pevida, M.G. Plaza, F.

Rubiera, J.J. Pis, High-pressure gasification reactivity of biomass chars produced at

different temperatures, Journal of Analytical and Applied Pyrolysis 85 (2009) 287-

293.

39. J.M. Encinar, F.J. Beltrán, A. Ramiro, J.F. González, A. Bernalte, Combustion

kinetics of agricultural wastes, Journal of Chemical Technology and Biotechnology

64 (1995) 181-187.

24

40. M.V. Gil, J. Riaza, L. Álvarez, C. Pevida, J.J. Pis, F. Rubiera, Kinetic models for the

oxy-fuel combustion of coal and coal/biomass blends chars obtained in N2 and CO2

atmospheres, Energy 48 (2012) 510-518.

41. T.F. Smith, Z.F. Shen, J.N. Friedman, Evaluation of coefficients for the weighted

sum of gray gases mode, Journal of Heat Transfer-Transmission ASME 104 (1982)

602-608.

42. C. Yin, Nongray-gas effects in modeling of large scale oxy-fuel combustion

processes, Energy and Fuels 26 (2012) 3349-3356.

43. P. Glarborg, A.D. Jensen, J.E. Johnsson, Fuel nitrogen conversion in solid fuel fired

systems, Progress in Energy and Combustion Science 29 (2003) 89-113.

44. L. Álvarez, M. Gharebaghi, J.M. Jones, M. Pourkashanian, A. Williams,J. Riaza, C.

Pevida, J.J. Pis, F. Rubiera, CFD modelling of oxy-coal combustion: Prediction of

burnout, volatile and NO precursors release, Applied Energy 104 (2013) 653-665.

45. J.M. Jones, M. Pourkashania, A. Williams, L. Rowlands, Q. Zhu, K.M. Thomas,

Conversion of char nitrogen to NO during combustion, Journal of the Energy

Institute 77 (2004) 82-89.

46. P. Molcan, G. Lu, T. Le Bris, Y. Yan, B. Taupin, S. Caillat, Characterisation of

biomass and coal co-firing on a 3 MWth Combustion Test Facility using flame

imaging and gas/ash sampling techniques, Fuel 88 (2009) 2328-34.

25

Table 1. Proximate and ultimate analyses of the samples used in this work

Sample OW HVN SAB Origin Spain Spain S. Africa Rank sa hvb Proximate analysis (wt %)Moisture 9.0 1.1 2.4 Volatile matter (db) 71.9 9.2 29.9 Ash (db) 7.6 10.7 15.0 Fixed carbona (db) 20.5 80.1 55.1 Ultimate analysis (wt %, db)C 50.2 81.9 69.3 H 6.1 3.1 4.2 N 1.8 1.7 1.8 S 0.2 1.4 0.8 Oa 34.1 1.2 8.9 High heating value (MJ/kg, db) 19.9 31.8 27.8 sa: semi-anthracite; hvb: high-volatile bituminous coal.

db: dry basis.

a Calculated by difference.

26

Table 2. Inputs of the CFD code for the gases and coal and biomass blends feed rates

Atmosphere Gas inlet

(g/min)

Coal and biomass blends mass flow rate (g/min)

HVN-OW blends SAB-OW blends

0%OW 10%OW 20%OW 0%OW 10%OW 20%OW

21%O2/79%N2 1.548 0.110 0.114 0.119 0.105 0.135 0.139

21%O2/79%CO2 2.118 0.110 0.114 0.119 0.105 0.135 0.139

30%O2/70%CO2 2.058 0.157 0.164 0.171 0.147 0.194 0.196

35%O2/65%CO2 2.016 0.182 0.190 0.198 0.175 0.225 0.231

27

Table 3. Devolatilisation data inputs for the CFD code

Sample A (s-1) Ead (kJ mol-1)

OW 1.9×109 127.0

HVN 3.60×1014 229.7

SAB 4.68×1011 155.9

28

Table 4. Char combustion kinetic parameters employed

Coal Char obtained in N2 Char obtained in CO2

A (s-1) Ead (kJ mol-1) A (s-1) Ead (kJ mol-1)

OW 1×108 89.4 - -

HVN 5.09×104 127 8.10×103 117

SAB 9.48×104 120 2.31×105 125

29

Table 5. Nitrogen content (%mass) in char and volatiles after coal devolatilisation in the EFR in N2 and CO2

Coal Char obtained in N2 Char obtained in CO2

N-char N-volatile N-char N-volatile

OW 2.25 1.57 2.11 1.67

HVN 1.93 1.94 1.90 2.31

SAB 2.38 2.64 2.26 2.70

30

Table 6. Experimental and predicted burnouts and NO emissions for HVN-OW and SAB-OW combustion in air and oxy-fuel atmospheres (21-35% O2)

Burnout

(%)

21%O2/79%N2 21%O2/79%CO2 30%O2/70%CO2 35%O2/65%CO2

Exp Pred Exp Pred Exp Pred Exp Pred

HVN 79.5 79.5 77.1 77.1 81.0 80.8 82.9 82.3

90HVN-10OW 80.5 80.6 79.5 78.5 83.3 84.3 85.4 86.7

80HVN-20OW 83.0 84.0 81.1 81.0 83.9 84.0 85.9 85.6

SAB 92.5 92.3 90.2 89.7 93.9 92.7 94.7 94.6

90SAB-10OW 93.9 93.5 92.7 91.4 95.0 94.9 95.7 95.0

80SAB-20OW 95.2 94.3 93.0 93.1 95.8 95.6 97.8 95.7

NO emissions

(ppm)

21%O2/79%N2 21%O2/79%CO2 30%O2/70%CO2 35%O2/65%CO2

Exp Pred Exp Pred Exp Pred Exp Pred

HVN 384 390 360 362 578 573 626 613

90HVN-10OW 392 398 346 350 547 542 633 585

80HVN-20OW 395 390 339 343 548 551 633 578

SAB 400 388 359 360 498 504 474 527

90SAB-10OW 333 346 290 295 418 430 436 422

80SAB-20OW 270 293 181 237 330 340 350 354

31

Figure captions

Fig. 1. Schematic diagram of the entrained flow reactor (EFR)

Fig. 2. Predicted temperature (K) in the EFR when co-firing HVN-OW in (a)

21%O2/79%N2, (b) 21%O2/79%CO2, (c) 30%O2/70%CO2 and (d) 35%O2/65%CO2.

Length scale is 140 cm and 40 cm respectively.

Fig. 3. Predicted area-weighted average temperature during combustion in air and

oxy-fuel environments of HVN-OW blends.

oxy-fuel atmospheres

Fig. 4. Predicted temperature (K) in the EFR when co-firing SAB-OW in (a)



Fig. 5. Predicted area-weighted temperature during combustion in air and oxy-fuel

environments of blends SAB-OW.

Fig. 6. Predicted burning rate (kg/s) in the EFR when co-firing HVN-OW in (a)



Fig. 7. Predicted burning rate (kg/s) in the EFR when co-firing SAB-OW in (a)



Fig. 8. Predicted NO concentration (ppm) in the EFR when co-firing HVN-OW in (a)



Fig. 9. Predicted NO concentration (ppm9 in the EFR when co-firing SAB-OW in (a)



32

N2

Gas cylinders

Cooled injector

Feeding system

Samplingprobe

Cyclone

O2

CO

CO2

Mass flowcontrollers

Gas analysers

To vent

Filter

Pre -heater

Reaction tube

O2 CO2

NO

SO2

N2

Gas cylinders

Cooled injector

Feeding system

Samplingprobe

Cyclone

O2

CO

CO2

Mass flowcontrollers

Gas analysers

To vent

Filter

Pre -heater

Reaction tube

O2 CO2

NO

SO2

Figure 1. Schematic diagram of the entrained flow reactor (EFR).

33

80HVN

Tempe

rature (K

)

HVN 90HVN

a) b)

Tempe

rature (K

)HVN 90HVN 80HVN

Tempe

rature (K

)

HVN

c)

90HVN 80HVN 90HVN 80HVN

d)

HVN

Tempe

rature (K

)80HVN

Tempe

rature (K

)

HVN 90HVN

a)

80HVN

Tempe

rature (K

)

HVN 90HVN

a)Tempe

rature (K

)

HVN 90HVN

Tempe

rature (K

)

HVN HVN HVN 90HVN

a) b)

Tempe

rature (K

)HVN 90HVN 80HVN

b)

Tempe

rature (K

)HVN 90HVN 80HVN

Tempe

rature (K

)HVN 90HVN 80HVNHVN 90HVNHVN 90HVN 80HVN

Tempe

rature (K

)

HVN

c)

90HVN 80HVN

Tempe

rature (K

)

HVN

c)

90HVN

Tempe

rature (K

)

HVN

c)

Tempe

rature (K

)

HVN

Tempe

rature (K

)

HVN

c)

90HVN90HVN 80HVN 90HVN 80HVN

d)

HVN

Tempe

rature (K

)

90HVN 80HVN

d)

HVN

Tempe

rature (K

)d)

HVN

Tempe

rature (K

)

HVNHVN

Tempe

rature (K

)

Figure 2. Predicted temperature (K) in the EFR when co-firing HVN-OW in (a)


Length scale is 140 cm and 40 cm respectively

34

21%O2/79%N2

1000

1100

1200

1300

1400

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

Distance (m)

Tem

pera

ture

(K)

HVN

90HVN-10OW

80HVN-20OW

21%O2/79%CO2

1000

1100

1200

1300

1400

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

Distance (m)Te

mpe

ratu

re (K

)

HVN

90HVN-10OW

80HVN-20OW

30%O2/70%CO2

1000

1100

1200

1300

1400

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

Distance (m)

Tem

pera

ture

(K)

HVN

90HVN-10OW

80HVN-20OW

35%O2/65%CO2

1000

1100

1200

1300

1400

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

Distance (m)

Tem

pera

ture

(K)

HVN

90HVN-10OW

80HVN-20OW

Figure 3. Predicted area-weighted temperature during combustion in air and oxy-fuel

environments of HVN-OW blends.

35

Tempe

rature (K

)

SAB

Tempe

rature (K

)

b)

90SAB 80SAB

a)

90SABSAB

Tempe

rature (K

)

80SAB

SAB

c)

90SAB 80SAB 80SABSAB

Tempe

rature (K

)d)

90SAB

Tempe

rature (K

)

SAB

Tempe

rature (K

)

b)

90SAB 80SAB

a)

90SABSAB

Tempe

rature (K

)

80SAB

SAB

c)


Tempe

rature (K

)d)

90SAB

SAB

Tempe

rature (K

)

b)

90SAB 80SAB

a)

90SABSAB

Tempe

rature (K

)

80SAB

SAB

c)

90SAB 80SAB

SAB

Tempe

rature (K

)

b)

90SAB 80SABSAB

Tempe

rature (K

)

b)

90SABSAB

Tempe

rature (K

)

b)

SAB

Tempe

rature (K

)

b)

SAB

Tempe

rature (K

)

SAB

Tempe

rature (K

)

b)

90SAB 80SAB

a)

90SABSAB

Tempe

rature (K

)

80SAB

SAB

c)

90SAB 80SAB

a)

90SABSAB

Tempe

rature (K

)

80SAB

a)

90SABSAB

Tempe

rature (K

)

80SAB90SABSAB

Tempe

rature (K

)

80SABSAB

Tempe

rature (K

)

SAB

Tempe

rature (K

)

80SAB80SAB

SAB

c)

90SAB 80SABSAB

c)

90SABSAB

c)

SAB

c)c)


Tempe

rature (K

)d)

90SAB 80SABSAB

Tempe

rature (K

)d)

90SABSAB

Tempe

rature (K

)d)

90SAB

Tempe

rature (K

)d)

90SAB

Figure 4. Predicted temperature (K) in the EFR when co-firing SAB-OW in (a)



36

21%O2/79%N2

1000

1100

1200

1300

1400

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

Distance (m)

Tem

pera

ture

(K)

SAB

90SAB-10OW

80SAB-20OW

21%O2/79%CO2

1000

1100

1200

1300

1400

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

Distance (m)

Tem

pera

ture

(K)

SAB

90SAB-10OW

80SAB-20OW

30%O2/70%CO2

1000

1100

1200

1300

1400

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

Distance (m)

Tem

pera

ture

(K)

SAB

90SAB-10OW

80SAB-20OW

35%O2/65%CO2

1000

1100

1200

1300

1400

0.0 0.2 0.4 0.6 0.8 1.0 1.2 1.4

Distance (m)

Tem

pera

ture

(K)

SAB

90SAB-10OW

80SAB-20OW

Figure 5. Predicted area-weighted temperature during combustion in air and oxy-fuel

environments of blends SAB-OW.

37

Burning rate (kg/s)

HVN

b)

90HVN 80HVN

Burning rate (kg/s)

HVN

a)

80HVN90HVN

HVN

Burning rate (kg/s)

c)

80HVN90HVN HVN

Burning rate (kg/s)

d)

80HVN90HVN

Burning rate (kg/s)

HVN

b)

90HVN 80HVN

Burning rate (kg/s)

HVN

b)

90HVN

Burning rate (kg/s)

HVN

b)

Burning rate (kg/s)

HVN

b)

HVNHVN

b)

90HVN 80HVN

Burning rate (kg/s)

HVN

a)

80HVN90HVN

Burning rate (kg/s)

HVN

a)

Burning rate (kg/s)

HVN

a)

HVN

a)

80HVN80HVN90HVN90HVN

HVN

Burning rate (kg/s)

c)

80HVN90HVNHVN

Burning rate (kg/s)

c)

HVN

Burning rate (kg/s)

HVNHVN

Burning rate (kg/s)

c)

80HVN80HVN90HVN90HVN HVN

Burning rate (kg/s)

d)

80HVN90HVNHVN

Burning rate (kg/s)

d)

HVN

Burning rate (kg/s)

HVNHVN

Burning rate (kg/s)

d)

80HVN80HVN90HVN

Figure 6. Predicted burning rate (kg/s) in the EFR when co-firing HVN-OW in (a)



38

SAB 80SAB

d)

Burning rate (kg/s)

90SAB80SABSAB

Burning rate (kg/s)

c)

90SAB

80SABSAB

Burning rate (kg/s)

a)

90SABBu

rning rate (kg/s)

80SABSAB

b)

90SAB

SAB 80SAB

d)

Burning rate (kg/s)

90SABSAB 80SAB

d)

Burning rate (kg/s)

90SAB 80SAB

d)

Burning rate (kg/s)

90SAB

d)

Burning rate (kg/s)

90SAB

d)

Burning rate (kg/s)

d)

Burning rate (kg/s)

90SAB80SABSAB

Burning rate (kg/s)

c)

90SAB 80SAB80SABSAB

Burning rate (kg/s)

c)

90SABSAB

Burning rate (kg/s)

c)

SAB

Burning rate (kg/s)

c)

90SAB90SAB

80SABSAB

Burning rate (kg/s)

a)

90SAB 80SABSAB

Burning rate (kg/s)

a)

90SABSAB

Burning rate (kg/s)

a)

90SABSAB

Burning rate (kg/s)

a)

SAB

Burning rate (kg/s)

SAB

Burning rate (kg/s)

a)

90SABBu

rning rate (kg/s)

80SABSAB

b)

90SABBu

rning rate (kg/s)

80SABSAB

b)

90SAB 80SABSAB

b)

90SABSAB

b)

90SABSAB

b)

90SABSAB

b)

SABSAB

b)

90SAB

Figure 7. Predicted burning rate (kg/s) in the EFR when co-firing SAB-OW in (a)



39

80HVN90HVNHVN

NO (p

pm)

d)

HVN

NO (p

pm)

a)

80HVN HVNNO (p

pm)

b)

90HVN 80HVN

HVN

NO (p

pm)

c)

90HVN80HVN

90HVN

80HVN90HVNHVN

NO (p

pm)

d)

HVN

NO (p

pm)

a)

80HVN HVNNO (p

pm)

b)

90HVN 80HVN

HVN

NO (p

pm)

c)

90HVN80HVN 80HVN90HVNHVN

NO (p

pm)

d)

80HVN90HVNHVN

NO (p

pm)

d)

90HVNHVN

NO (p

pm)

d)

HVN

NO (p

pm)

HVNHVN

NO (p

pm)

d)

HVN

NO (p

pm)

a)

80HVNHVN

NO (p

pm)

a)

HVN

NO (p

pm)

a)

HVN

NO (p

pm)

HVNHVN

NO (p

pm)

a)

80HVN HVNNO (p

pm)

b)

90HVN 80HVNHVNNO (p

pm)

b)

90HVNHVNNO (p

pm)

b)

HVNNO (p

pm)

HVNHVNNO (p

pm)

b)

90HVN 80HVN

HVN

NO (p

pm)

c)

90HVN80HVNHVN

NO (p

pm)

c)

90HVNHVN

NO (p

pm)

c)

HVN

NO (p

pm)

HVNHVN

NO (p

pm)

c)

90HVN80HVN

90HVN

Figure 8. Predicted NO concentration (ppm) in the EFR when co-firing HVN-OW in

(a) 21%O2/79%N2, (b) 21%O2/79%CO2, (c) 30%O2/70%CO2 and (d) 35%O2/65%CO2.


40

NO (p

pm)

SAB

NO (p

pm)

d)

90SAB 80SAB

SAB

NO (p

pm)

b)

90SAB 80SAB

a)

SAB 90SAB 80SAB

80SABSAB

NO (p

pm)

c)

90SAB

NO (p

pm)

SAB

NO (p

pm)

d)

90SAB 80SAB

SAB

NO (p

pm)

b)

90SAB 80SAB

a)

SAB 90SAB 80SAB

80SABSAB

NO (p

pm)

c)

90SAB SAB

NO (p

pm)

d)

90SAB 80SABSAB

NO (p

pm)

d)

90SABSAB

NO (p

pm)

d)

SAB

NO (p

pm)

SABSAB

NO (p

pm)

d)

90SAB 80SAB

SAB

NO (p

pm)

b)

90SAB 80SABSAB

NO (p

pm)

b)

90SABSAB

NO (p

pm)

b)

SAB

NO (p

pm)

SAB

NO (p

pm)

b)

90SAB 80SAB

a)

SAB 90SAB 80SAB

80SABSAB

NO (p

pm)

c)

90SAB

a)

SAB 90SAB 80SAB

a)

SAB 90SAB

a)a)

SAB 90SAB 80SAB

80SABSAB

NO (p

pm)

c)

90SAB 80SABSAB

NO (p

pm)

c)

90SABSAB

NO (p

pm)

c)

SAB

NO (p

pm)

SABSAB

NO (p

pm)

c)

90SAB

Figure 9. Predicted NO concentration (ppm) in the EFR when co-firing SAB-OW in

(a) 21%O2/79%N2, (b) 21%O2/79%CO2, (c) 30%O2/70%CO2 and (d) 35%O2/65%CO2.


biomass co-firing under oxy-fuel conditions: a computational fluid...

Documents